We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:


References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Do Answers to Boolean Questions Need Explanations? Yes

Abstract: Existing datasets that contain boolean questions, such as BoolQ and TYDI QA , provide the user with a YES/NO response to the question. However, a one word response is not sufficient for an explainable system. We promote explainability by releasing a new set of annotations marking the evidence in existing TyDi QA and BoolQ datasets. We show that our annotations can be used to train a model that extracts improved evidence spans compared to models that rely on existing resources. We confirm our findings with a user study which shows that our extracted evidence spans enhance the user experience. We also provide further insight into the challenges of answering boolean questions, such as passages containing conflicting YES and NO answers, and varying degrees of relevance of the predicted evidence.
Comments: 9 pages
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2112.07772 [cs.CL]
  (or arXiv:2112.07772v1 [cs.CL] for this version)

Submission history

From: Sara Rosenthal [view email]
[v1] Tue, 14 Dec 2021 22:40:28 GMT (2258kb,D)

Link back to: arXiv, form interface, contact.